We are looking for driven and talented students to push the boundaries of research in the following areas:

• Underwater computational imaging. The advent of light field cameras offers exiting opportunities to design novel imaging systems tailored to the challenges of using light underwater: wavelength and range-dependent attenuation, backscatter and significant refraction.

• Multiscale classification of visual composites of reefs. Our ability to create high resolution, large extent (much larger than a single image footprint) mosaics and 3D reconstructions opens the possibility of image understanding at mulitple spatial scales and semantic resolutions.

• Real time exploration and mapping with vision and sonar. Move away from traditional preprogrammed paths to adaptively planning the robot’s path dependent on observations gathered so far, using appropriate SLAM algorithms to build the map and decide where to sample next.

• Multimodal habitat sampling and prediction. Fully covering the seafloor with images is unfeasible. Instead, we use lower resolution and less discriminative acoustics to cover broad areas and selectively sample with optical images to interpret the acoustics. We seek to build representations that allow predictions of visual information and can be used to plan further sampling.

• Scalable autonomous systems for characterisation of large scale marine protected areas. When characterising benthic habitats of hundreds or thousands of square km, our current approach relying on manned vessels and AUVs with endurances of hours are too expensive. We are investigating the use of autonomous surface vessels supporting simple imaging platforms for long, untended deployments.

• Automated cooperative, risk-aware planning for surveys with heterogenous platforms. The use of multiple, complementary autonomous platforms increases the complexity of planning and executing missions. Methods that provide robust and sensible planning are key to enable regular operation of robotic systems consisting of multiple platforms.

Information of the ACFR’s graduate program, admissions and scholarships can be found here.

If you’d like more information on these projects, contact stefanw@acfr.usyd.edu.au or o.pizarro@acfr.usyd.edu.au